Epidemiologic studies of osteoarthritis have used x-rays as the only modality to image pathology and to define disease. X-rays reflect advanced bony changes of osteoarthritis about cartilage loss but provide no information about intraarticular soft tissue pathology, cartilage loss of bone marrow lesions. The source of knee symptoms is currently unknown but soft tissue and bone marrow lesions are likely causes. Further many persons in the population with knee symptoms do not have radiographic OA, but they may OA if examined by a more accurate and sensitive tool, magnetic resonance imaging (MRI). MRI accurately images cartilage, bone and soft tissue but has not been used to evaluate persons in the general population nor to evaluate those with knee symptoms. This study has two overall specific aims: 1) using longitudinal data from the Framingham Offspring Study to examine the effect of specific forms of physical activity, quadriceps weakness and of vitamin D and PTH on the development or progression of radiographic osteoarthritis; and 2) in a cross-sectional evaluation, to evaluate the prevalence of knee MRI findings in a community-based sample and to evaluate the correlation of these findings with knee symptoms. To examine these questions, the investigators will assess knee and hand OA in 2300 subjects; 1510 of these are members of the Framingham Offspring Cohort (these will be the subjects in the longitudinal study), who had a baseline examination for knee and hand OA in 1993-1994. At the time of that previous baseline examination, physical activity was evaluated, isometric quadriceps strength and hand and knee symptoms assessed and radiographs obtained. In addition, they will recruit 790 adults age 50 and over who live in Framingham or surrounding towns who will compromise the remainder of the sample and will permit, the investigators state, drawing a population based sample diverse in racial background. Subjects will undergo a comprehensive examination of knee and hand symptoms, imaging and risk factors for disease. The investigators state that this will be the first population-based study to include MRI and it will be the first to explore the complex relationship between specific physical activities and knee OA. They also state that in its cross-sectional focus on frequent knee symptoms, it will be the first to begin to evaluate using epidemiologic methods to study knee symptoms, the entity that has the most public health and clinical importance.